diff --git a/src/llama-context.cpp b/src/llama-context.cpp
index 3cc27717ec..b28726622a 100644
--- a/src/llama-context.cpp
+++ b/src/llama-context.cpp
@@ -3688,6 +3688,9 @@ llama_context * llama_init_from_model(
             LLAMA_LOG_ERROR("%s: SPLIT_MODE_TENSOR requires flash_attn to be enabled\n", __func__);
             return nullptr;
         }
+        if (model->get_split_state_ud.n_devices == 1) {
+            LLAMA_LOG_WARN("%s: SPLIT_MODE_TENSOR being used for a single device is not recommended\n", __func__);
+        }
     }
 
     if ((model->hparams.is_mla() || model->arch == LLM_ARCH_DEEPSEEK4) && params.type_k != params.type_v) {